Tories and Labour at 35%

March's Populus poll for The Times puts both main parties at 35%.  The LibDems are up two points to 20%.  This is in line with other polls which suggest that Britain's third party has recovered its balance and is back in political business.

Peter Riddell writes about "the silly talk" that "the Cameron effect [is] stalling":

"The monthly shifts mean little. The key point is that the Tory vote share is now 5.7 points higher on average than over the previous six months, more than his predecessors achieved. The main improvement has been among professionals and managers, where support has risen from 34 to 42 per cent."

Contrary to other polls which suggest that Gordon Brown would boost Labour, Populus find that Cameron would do better against Mr Brown:

"When half the sample were asked how they would vote at the next election, 40 per cent said the Tories under Mr Cameron, 34 per cent Labour under Mr Brown and 20 per cent the Lib Dems under Sir Menzies, with others at 6 per cent."

The poll also suggests that David Cameron is wise to let newspapers do the running on Jowell-gate.  Populus found that just "15 per cent think the Labour Government is more guilty of sleaze than the previous Tory one, with 11 per cent thinking it less sleazy."  Two-thirds think both parties are as sleazy as each other.

Comments

[Jonathan Sheppard]

Any one suggest why the Lib Dems are so high (in my view) at 20% in the polls?

[Frank Young]

It might has something to do with a large number of people undecided or unsure about the two main parties.

[Jonathan Sheppard]

I just cant imagine their leadership election created any of the positive momentum that the leaderhip election in the Conservative party did.

[Ted]

LDs have been running at around 20% (occassionally up to 24/25%, in Dec/Jan down to 16% or so) for most of the time since the last election. It's not a leadership bounce rather a leadership no effect. They did suffer just after Charlie/Oaten/Hughes affairs but these seem to have been viewed as irrelevancies by LD voters (wish our Tory scandals had ben so easily forgiven - perhaps they don't matter as LD support doesn't expect power?)

[Annabel Herriott]

Doesnt it depend on who the pollsters actually contact to poll? I have never been polled by one of these organisations in my entire life. So to extrapolate, The times is left wing, so calls up a load of lefties, some of which are obliged to take notice of the obvious, ie Tories are coming up in the world. The rest are so tribal, they are like Comical Ali. Remember him? I still have a chuckle at the dear soul, grimly talking up Saddam, when the U.S. troops can be clearly seen behind him.

[Editor]

Annabel: There is absolutely no evidence that Populus or other pollsters are ideologically biased in that they call up a slanted selection of people. I think it is fair to question the nature of questions that some pollsters ask but not their professional integrity. I have to insist that you do not repeat such suggestions on this blog.

[Ted]

Interesting to compare Populus & YouGov since election.
Populus Tory lead/deficit
Jun na Jul -12 aug na sep -2 oct -10 nov -8 dec -3 jan-3 feb +1
YouGov Tory lead/deficit
Jun -7 Jul -5 aug -9 sep -8 oct -8 nov -2 dec +2 jan-1 feb +2

There was a very limited post election bounce for Labour - a couple of percent - but a noted post election fall for us until DCs election brought Tories level with Labour.

However even with the Gordon v Dave 34%:40% outturn we could still not have a majority on Baxter even allowing for boundary changes. It means we must put a lot of effort into the marginals where we can rectify the Labour bias by better campaigning.

[Annabel Herriott]

Not knowing anything about polling, then why do I have this strong perception, Editor. The comment was made in good faith.How then, do people get chosen to be polled? I repeat I have never been polled in all my 71 years, 45 of it spent in the NHS. One would have thought I would have had a passing chance?

[Jonathan Sheppard]

Do people think that the electorate is much more sophisticated than it use dto be? What would it take for the Tories to get up to 40% or should we not really worryto much - as why do we need to be at that level of support years before an election?

[Frank Young]

Annabel it is unlikely. Polls tend to be based on 1000 ish participants which are then adjusted to make allowances for different groups in society so the polls are broadly representative within a margin of error.

There are 45 000 000 million voters!! I'm only 23 and I’ve been asked before. Just luck I’m afraid. If you want to be included in a poll go on the yougov website and you can involve yourself there.

[Ted]

Oppositions are often squeezed by the "lets not let go of Nanny's hand" tendency so its usually better to have a really good lead just before election campaign. There can be upsets (as on that glorious day in 1970) but its good for the opposition to come in to election as the likely winners.

Also as we experienced from 1992-1997 being behind, losing locals & by-elections saps the party morale. Gordon's a clever politician but having been viewed for so long as the great hope I don't know how the Labour party would react if his leadership resulted in poor poll ratings.

Someone on politicalbetting referred to this just being Labour mid term blues - barely 9 months after the election. Hopefully by next year we'll start to see the real mid-term blues (Cons 41%, LDs 26%, Lab 25% = tory landslide...well you can dream)

[Chris Palmer]

"Any one suggest why the Lib Dems are so high (in my view) at 20% in the polls?" - Jonathan Sheppard

Because those who are likely to vote Lib Dem don't care about the scandelous activities they got up to. Also, I saw your contribution to the Tessa Jowell Your Views on the Telegraph. Very good!

[kingbongo]

Editor

Although Annabel is wrong to question the integrity of polling organisations she is right to identify a Labour bias in polls compared to actual outcomes. YouGov seem to have the best overall predicitve ability so I have faith in them but I believe despite their best efforts there was still a bias to Labour of about 3% at the last election compared to the actual result. I think MORI and Populus are the two least useful pollsters but I'll take a lead for DC in the DC v GB polls from anyone and will choose to concentrate on that!!

[Annabel Herriott]

Thank you Frank, and Kingbongo. I am not usually wrong when I get that female intuition! (Now I am really in trouble) I shall get the yougov website up, once I have finished a batch of jam for our Hospice. Isntthat what we wimmen are supposed to be doing???

[Sam Coates]

Liam Fox was giving interesting briefings at CCHQ last year, he looked at all of the elections in the last 50 years or so and pointed out how polls consistently over-estimate Labour support by 3 percentage points at the very least, whilst substantially under-estimating Conservative (& LibDem) support.
It's not a conspiracy, just a methodological bias that we should be aware of.

[Andrew Cooper]

The latest Populus poll for The Times has prompted various comments and questions about how polls are conducted - and how accurate they are, so I thought the following may be helpful.

Populus conducts its polls using random digit dialling telephone fieldwork. This works as follows. The software running the call centres draws random phone numbers from a database of every domestic telephone directory in the country. Numbers are drawn proportionately from each region so that the overal sample has the right regional composition. The final digit of each telephone number so chosen is randomly changed so as to get to unlisted numbers. What this all means is that everyone in Great Britain with a fixed line telephone has an equal random chance of being polled. Data collected in this way has, according to the laws of statistics, a 95% chance of being accurate to plus or minus 2.5% (for the 1,500 samples we do in polls for The Times): if it were possible to interview everyone in Britain with a fixed line telephone (which obviously it isn't) their views should be within 2.5% either side of what the poll reports. ICM use the same fieldwork methods. So do NOP. MORI often do - though sometimes continue to interview face-to-face. YouGov are of course unique as an internet-only polling company.

The polls at the last election were more accurate than they have ever been before. The most accurate was NOP, polling for The Independent, who got each party's vote share exactly right. Populus was in fact next most accurate: our eve of election poll conducted for Lord Ashcroft and published in his book 'Smell the coffee: a wake-up call to the Conservative Party' got the Labour and Lib Dem vote shares exactly right but understated the Conservative vote by 1%.

It is true to say that, though polls are much more accurate than in the past (for easily definable reasons) they still have a tendency slightly to overstate Labour's vote share relative to the Conservatives. We constantly work to refine and improve our methodology to make our polls more accurate and continue to do so in this regard. The evidence is that the main cause of this slight tendency to overstate Labour's support is that Labour supporters who tell pollsters they are absolutely certain to turn out and vote are, on recent evidence, fractionally less likely actually to do so than Conservative supporters who say they are certain to vote. This is very hard to adjust for in a methodologically robust and consistent way, though we continue to look for ways to address it.

Polls are not precision-measuring devices, they are fairly blunt instruments. But, as such they have a good record of accuracy: all the main polling companies were well within the statistical margin of error at the last election.

[Sean Fear]

The eve of polls at the last election were:-

Yougov; Con 32%, Lab 37%
Populus Con 32%, Lab 38%
ICM; Con 32%, Lab 38%
MORI; Con 33%, Lab 38%
NOP; Con 33%, Lab 36%.

The outcome was Con 33%, Lab 36% - so the bias to Labour was very slight.

That's not to deny that polls generally *do* overstate Labour's support (by quite substantial margins in 1992, 1997, and 2001). In the last two of those, Labour's victory was so emphatic that nobody but the pollsters noticed that their lead was exaggerated.

[Peter Harrison]

At the last election most of the major polling organisations were within the margin of error in their final predictions and NOP were spot on. MORI also produced a reasonable final poll although their polls during the campaign were all over the place.

The pollsters have put in a lot of effort to reduce the Labour bias in their results, which seems to be mainly due to a couple of factors:

- Conservative voters are more likely to refuse to answer questions on their voting intentions
- the poor image of the party in recent years has meant that some people will lie to pollsters and claim to vote for another party even though they really intend to vote Conservative

There have only been two occasions where the pollsters got the result of a General Election badly wrong - 1970 and 1992. The 1970 error seems to have been caused by a combination of a late swing to the Conservatives plus the fact that most pollsters at that time did not take into account whether or not the voter actually intended to vote (only one did and they called the election correctly). The problems in 1992 have generally been blamed on the factors listed above.

I don't think there is any evidence of a conspiracy. There is evidence that there was a methodological bias but there is also evidence that the pollsters have been working to correct that and they seem to be getting there, although they aren't quite there yet.

[Sean Fear]

But I would reiterate that no polling company has any vested interest in producing anything other than accurate information . They get most of their income from market research, and only a tiny proportion from political polls. Hence, it's very much in their interests to ensure that the political polls (which get the most publicity) are accurate.

[Anthony Wells]

Populus are hardly a bunch of biased lefties - they are run by two former senior CCO staffers (not that they are biased righties either - they are professional pollsters).

For phone polls about 1000 people are selected using randomised telephone numbers, so assuming you have a telephone line you are as likely to be polled as anyone else (if you don't have a phone, you are excluded - one of the problems facing phone polling). Out of an adult population of 40 million odd, the chances of you being picked for any particular poll though are vanishingly small.

[Ted]

Andrew Cooper

Very good to have Populus itself responding to the comments on this blog, thank you for your thorough explanation.

[Ted]

Populus & Anthony Wells posting! - I'll have to put a lot more thought into comments about polls....

[Anon]

I actually work for one of the "big three" polling companies.

Whenever a client commissions a poll there are a number of stages. Clients will sometimes attempt to influence the outcome by requesting leading questions. However any reputable polling company would not allow this to happen, as too many rogue results would harm its reputation.

No client, (political party, newspaper or otherwise), would be allowed to influence the methodology or sampling procedure.

There are factors that can influence the outcome, but we make every effort to eliminate these. The most usual problem is bias on the part of the interviewer.

A typical national poll would have 1,800 respondents. Of these, 1,200 or so would be used, the balance being discarded as they are "out of quota" in terms of social grading or other factors.

A random sample of 30% of all respondents are back-checked by phone or letter (or in some cases personal visit). If any back-check highlights inaccuracies, all of that Interviewers work is back-checked and possibly discarded from the analysis.

Likewise, any abnormal results (ie, results from a particular Acorn Sample varying <5%> from the national average) are also investigated for accuracy.

As stated be a previous post - it is not in our interests to produce biased or misleading data.

Footnote: I am a regular user of this site but I have posted this anon for reasons I hope users will appreciate.

[Stephen B]

Can someone remind me whether telephone polling has to take into account TPS or does that just apply to calls where people try to sell you something?

[Anthony Wells]

No, TPS only applies to marketing and sales calls, not market research. (Though if a poll contained any element of marketing it would do - the MRS guidance is here http://www.mrs.org.uk/standards/tps.htm)

[William Norton]

I've always assumed that opinion polls have an in-built "winner's surplus", that don't knows/won't votes/waiverers tend to plump for whoever seems to be ahead and likely to win.

Polls have over-stated Labour during the 1990s and since because, bluntly, only they have looked like winning. On the other hand, I seem to recall that Thatcher never actually polled as high as her pre-election poll rating (perhaps someone will helpfully put me right on that?).

BTW - the Society of Pedants have asked me to correct Frank's earlier remark that "There are 45 000 000 million voters!!" No, there aren't (that figure is about 7,000 times the population of the planet: not even Prescott's going to give Martians a postal vote). There are, of course, only about 45 million voters in the UK.

[Gareth]

Not strictly about this poll, but now here else to post this! After a fortnight of canvassing, I'm quietly confident that Labour are going to have a poor night on May 4th. I've never known the Labour vote so soft. I suspect it's going to stay at home in pretty large numbers. Anyone else any feedback?

[Tory T]

Yes, it's very useful to have direct responses from pollsters - thanks guys.

The biggest problem faced by pollsters who work for newspapers is that journalists (supported by editors) sometimes willfully misrepresent poll results to get the desired headline. Both Populus and YouGov have suffered from this.

It's vital that papers pulling such stunts are immediately named and shamed.

[Anthony Wells]

On Mrs T - in 1983 MORI had the Tories at 46%, when they actually got 44%. In 1987 they had the Tories at 44% when they actually got 43% - that's probably not a pattern though. In their previous 2 polls MORI had the Tories at 41%, and ICM had them at 42% (plus, of course, they is the case of the 1992 polls which seriously underepresented the Tory vote).

The "winners surplus" isn't so much don't knows plumping for the leading party - although it has the same effect. The reason is actually people who support an unfashionable party can be embarrassed to admit it, the so-called "spiral of silence". It doesn't always work against the losing party of course - while people were reluctant to admit to voting Tory in the 1992 election (the "Shy Tory" syndrome), they were in fact the most popular party. Equally it's now thought that during the last Parliament people were more reluctant to admit voting Labour (what Andrew Cooper calls "Bashful Blairites") though Labour went on to win.

Either way, ICM and Populus now make adjustments to account for the "spiral of silence" while self-completion of YouGov surveys ameliorates the problem in their polls.

[Andrew]

I'd be very interested to see a poll comparing Brown and Cameron amongst English voters. Since the story of the next election could well be a Conservative majority in England, but a Labour government by virtue of Scottish and Welsh votes, the polling organisations should be tracking this.

[Annabel Herriott]

Well! Well! I did stir up a bit of a hornets nest. There must be a few more old Tories like me around, so that was a really useful discussion, even though I did get my wrists slapped! Have finished my jam, (blackberry) so if the editor wants a pot brought to Manchester as a peace offering, then let him say so!! He could even do a little fundraising for the blog, if he is having a stall. Scones and a choice of homemade jam! Or, is that too like the rubber chicken circuit.

[Jonathan Sheppard]

If Tim doesnt want it - you can send me some!

[Editor]

I don't know if I'll be going to Manchester, Annabel, but the prospect of your scones and jam may make the journey worthwhile...

[Ted]

Gareth

Noted your comment on May elections, hope you are right and we see Labour in third place! - did you see post earlier this month on
http://www.vote-2006.co.uk/
where you can discuss/comment on May elections

[matt wright]

The really important finding is that our share of the poll amongst professionals and managers is up considerably. This was the group that had switched off us and I saw this for myself canvassing last year. This is great news and is deirectly the result of Cameron's approach,

Matt

[R UK]

Internet polls at the last election showed Labour as crushed with around 22% against Lib Dem 23% with Cons on 35%. Internet polling samples were vast.

One explanation is that postal voting accounted for about one third of Labour's total vote at the last election. It was organised by the Parties and was said to be highly questionable until the media fell silent about it.

Many went to vote but found their vote mysteriously already used. including John Humphreys and Mariela Frostrup.

Times said that postal votes would be 4 million up from 2 million. In fact there were 6 million. Murdoch was backing the Tories until he realised the extent of the postal vote, when he swung back to labour.

Since then John Humphreys has been silenced by a campaign of intimidation. No news stories are run by anyone on postal voting fraud anywhere. I believe Labour indeed lost the election, and only 'won' by massive postal fraud.

Why the silence? It's deafening. As for the polls, they are hopeless merely providing what is expected of them - by those who pay them - ultimately the government can finish them too if it chooses...as it can any BBC exec, civil servant that whistleblows (Kelly) or any media that dares to tell the truth.

They only get away with it because no one dares to even imagine what is really going on.

[Giffin Lorimer]

Always difficult to understand the LibDem supporter. There seems to be a strong element of 'communautarianisme' in their preference for collective mechanisms rather than individualism. Hard to tell Labour and LibDem apart on many issues.

But there is always a gulf between supporters and voters in every party. It may be more instructive to debate the latter, since they are our target audience in the South. I'd say they seem to be very environmentally conscious, and very concerned about doing 'the right thing' by those less fortunate.

They are people who dislike the notion of 'the greater good' and who view unrestricted personal freedom and the dictatorship of the proletariat as being equally bad for the people in the margins of society.

They are people who like to think they vote altruistically - helping others before they help themselves [and thus the opposite of New Labour!]. Sleaze is therefore a really big issue to them as it's so, well, selfish. And war kills people so it's clearly a bad thing too. And they think that the rich should pay more taxes because they are richer than they are.

They aren't rich, and Black Wednesday frightened the bejayzus out of them, coming as it did after watching their neighbour's homes being repossessed during the early 1990's downturn. Brown hasn't done anything to make them want to vote Tory to kick him out of office. Mortgage rates are low and they feel rich because their house value is high.

The growing numbers of bankrupt, unemployed and economically inactive probably haven't affected them yet, since they work for big firms or a government agency, but with repossessions growing and their pensions being curtailed, that may change.

If we are to address their voter base, we have to tone down our 'the only good government is no government' rhetoric and talk about being in government to govern on behalf of everyone, especially the socially vulnerable and the environment. They probably don't care how we do it, so long as they don't have to suffer. Which is why we can't press for radical change, since that might cause them pain. But we will need credible policies and nice people to present them.

We also have to stir them out of their complacency about the economy to neutralise the "Nanny's hand" factor. But it still needs to be altruistically pitched, perhaps by showing how bad it is for young people and then frightening them about pensions. We also need to explain how it's better to encourage rich people to create jobs for the poor rather than giving money to them - Big Issue is a case in point. But unless the economy declines suddenly, that's probably a second-order issue for them.

[Sean Fear]

"Internet polls at the last election showed Labour as crushed with around 22% against Lib Dem 23% with Cons on 35%. Internet polling samples were vast"

No reputable poll gave a result remotely resembling those figures.

[Sean Fear]

"The really important finding is that our share of the poll amongst professionals and managers is up considerably."

You can't read too much into very small sub-samples within polls.

"This was the group that had switched off us and I saw this for myself canvassing last year. This is great news and is deirectly the result of Cameron's approach"

Well, it all depends where you were. In London and the South East, the Conservatives strengthened their vote among professionals and managers in 2005, although it certainly fell in the North and Midlands.

There seems to be quite a curious belief that this or that demographic group (professional people, 18-34 year old voters, women voters aged under 50 etc.) is somehow the key to winning the election. The fact is, the Conservatives need to improve their standing among *all* demographic groups to win.

[Ted]

Agree its about getting more in each demographic to vote for us but there is sense in targeting that demographic likely to bring best results in target seats.

What struck me in last election was that support for tories amongst homeowners with a mortgage was 30% to Labour 39%, and homeowners owning outright 43% to Labours 30%. Wondered if fear of the high mortgage rates was reason (and this would carry over into age groupings as well those with mortgage would be younger) or was it reflection of lower ratings among younger voters/ higher ratings among over 55s.

People probably fear higher mortgage rates more than they enjoy lower taxes - you can plan ahead for tax, mortgages can leap up in weeks.

[malcolm]

Sean,RUK seems to make this stuff up as he goes,this is not the first time he's made these allegations on this site.Much as I would like to believe them I think they're complete rubbish.

[Sean Fear]

Target seats come in many shapes and sizes. Targetting constituencies makes perfect electoral sense - but targetting particular social groups does not IMO.

[Edward]

Cameron's been, largely, loosening the vote. The real winning comes when policies, intelligently argued, begin to win over voters in the debates about them.

[Daniel Vince-Archer]

"Sean,RUK seems to make this stuff up as he goes,this is not the first time he's made these allegations on this site.Much as I would like to believe them I think they're complete rubbish."

I heard Lord Lucan was riding around on Shergar intercepting the postal voting forms as part of a campaign being co-ordinated by Elvis Presley in a UFO commissioned by Labour. Oh look, there goes a squadron of pigs zooming past my office window.

[R UK]

One Big Vote had a huge number of votes.

Malcolm - which fact is made up? I would like a chance to reply.

[R UK]

DVLA - you're drifting off again. Name one fact that you believe is wrong. I would be most grateful, if you could find a moment in amongst your creative writing.

[Anthony Wells]

Sigh...it's from this "poll" and it's worthless.

http://www1.onebigvote.co.uk/Results/ResultOver18.htm

The poll was wholly self-selecting and open access (i.e. if a party wanted to increase it's support on the poll it merely needed to email supporters and send them the link), it was possible to vote multiple times, there was no attempt to weight the results in any way to even the demographic make up of the UK, let alone the political make up.

A very, very large voodoo poll is still just a voodoo poll.

[malcolm]

RUK, Which poll had Labour on 22%?Which poll had a 'vast' sample size? When was 'Murdoch'(who I assume you mean Sun &Times)backing the Tories and how did he do so?

[R UK]

To argue that the world is round to a flat-earther is to shake his viewpoint so much that he feels uncomfortable. Next he ridicules the other's viewpoint.

Such a one as he will probably never even look at the evidence that the world might be round.

STATEMENT: Postal voting fraud at the last election was so large that it swung the result.

This statement is either right or it is wrong. Ridiculing it merely indicates that someone is not prepared to seriously consider the evidence.

Some people who stated that the earth was round met with violent deaths. Thank god for the internet.

Why is Blair allowed by the media to get away with his evil ways? And to me even more curious, why do his political opponents not even consider the possibility that he might be evil, and that he rigged the last general election?

And that we are living in a hall of mirrors where the truth was banned a long time ago. So effective has the banishment of truth been, that anyone who reintroduces it, is seen as of a lunatic fringe.

I'd prefer to be seen as that by a few flat-earthers than to be conned along with the rest.

[spagbob]

out of interest, exactly what percentage do we need to get over labour at the next election to have a majority?

[Richard]

I remember under William Hague when ITV teletext polls used to give the Tories 70%.

The Tories can't expect to do well in the polls without setting out clearly what they stand for and how they're going to get there. It's alright to talk about compassion etc but what will this involve?

[R UK]

Malcolm - go to your local library. get out the times, sun, sunday times, notw, and read them two weeks prior to the election.

They were all weighing in on Blair on postal vote fraud.

they stated that postal voting was going to be far bigger than previously.

they estimated 4 million. the papers seemed to be swinging towards supporting howard.

then a week later complete silence on postal voting fraud.

then the election. there were not 4 but 6 million postal votes. there was nil reporting about substantial fraud from any media. murdoch knew that it was time to keep quiet, but he tried to get the story into the public domain.

after reading these papers and seeing the pattern, you might be good enough to give me your views as to what explains it all.

[spagbob]

out of interest, exactly what percentage do we need to get over labour at the next election to have a majority?

[R UK]

many previously labour voters were puzzled by the election result.

many labour supporters abandoned blair over iraq...and many other topics.

canvassing on the ground showed a massive turning away from labour.

and yet labour got the votes it needed.

and yet people here seem quite happy about the 6 million postal votes - which were never investigated or reported on by any media.

labour only polled 9.4 million votes including all the postal ones they collected.

there is a story yet to be told. no media dares to tell it, or even ask a question. john humphreys started asking questions. look what happened to him.

[R UK]

if labour are able to manufacture postal votes - i have asked around a little bit and it is very easy for them to do it - then no conservative vote will be sufficient to win an election.

unless the postal voting system is dealt with, conservatives will never again win an election, given the other factors such as boundaries, and the 3 party split.

are you interested in asking any questions about postal voting fraud? If not, why not?

[malcolm]

What happened to John Humphreys RUK? Still waiting to know about the polls and when the Murdoch papers urged their readers to vote Conservative before the last election?

[Ted]

spagabob - it all depends on the Labour/Lib dem split + differential voting across country. We showed in last election that we could gain seats with little increase in our share of votes - including winning back LD seats while LD share went up. Lead probably needs to be around 7-8%.

Martin Baxters electoral calculus at
http://www.electoralcalculus.co.uk/
lets you try out various scenarios including a bit of tactical voting. Its not perfect but is fun as it lists the mps who'd lose their seats. Curently you need to add 11 to tories and subtract 11 from labour to roughly account for boundary changes.

Baxter does indicate that if we won 36% with Labour and Lib Dems at 28% (ie Labour lost 8% of its support we picked up 3% of that but rest went to LDs - unlikely but not that unlikely) we'd be close to majority,

With LDs at a more likely 23% of share, we'd need a 39% share to Labour's 30% to just gain majority.

[R UK]

papers - answered. two weeks prior to election. just get them and read them.

poll - answered. internet not controlled - but not necessarily wrong because of that.

john humphreys - they ran an intimidation campaign after the election to keep him pegged down. secretly recording his speech - remember? he tried to report on postal vote fraud and they came down on him like a ton of bricks. nothing mentioined now.

[Editor]

R UK - that's enough, thank you.

[James Hellyer]

I remember under William Hague when ITV teletext polls used to give the Tories 70%.

The problem is that those were "voodoo" polls. The sample base was self-selecting - only people who chose to go onto teletext and then phone in voted.

[Andrew]

The relationship between share of the vote and seats gained would alter pretty significantly if the Tories ever did get +9 on Labour. Such sites really only make sense with existing electoral geography, and tiny political changes at each election.

What I'd really like to see is polling heavily biased to the 100 most marginal seats - most of the rest are irrelevant, leaving voters there essentially disenfranchised. National polling is daft, but makes for nice simple headlines. After all, the parties target their resources to the crucial swing ultra-minorities, why don't the pollsters?

[spagbob]

thanks ted

[ted]

Andrew, Agree sites aren't perfect and I know its a long time since we won an election but 8% plus leads were what we got when we won elections. In 1992 we had an 8% lead, and a small 0.5% fall in our share of vote and only just got a majority. Putting 1992 shares through Baxters still only shows a slight variance from actual result

I take your point though on importance of the marginals - if we having any hope of winning next time with less than 7-8% lead we need to rectify the Labour imbalance in those seats (in 1992 had 1241 people not voted for us in marginals we'd have lost the majority)

[Richard]

"The problem is that those were "voodoo" polls. The sample base was self-selecting - only people who chose to go onto teletext and then phone in voted."

I know, I was just reminiscing. Those polls used to cheer me up in those dark days of Labour popularity, even though I knew they were nonsense. Just goes to show doublethink is possible!

[Richard]

"R UK - that's enough, thank you."

Clearly the editor is involved in the conspiracy too!

[James Hellyer]

Tim spent many a night in constituencies fabricating postal votes for the local Labour candidates. After the election he turned his attention to creating a forum for Conservative discord.

He's far more creative an anti-Conservative than the boy Miliband ;-)

[Ted]

slightly off thread - just seen in latest news that Sark is abandoning feudalism - the last hope for true conservatives dashed....
seems to be result of Human Rights lawyers

[Annabel Herriott]

My patch for my last eight years was !high ethnic" I can tell you lads categoricly that in the vast majority of households, DAD RULES OK!!! That meant, that even though Sayeeda had all the women in her pocket, when it came to postal votes, Shahid Malik got in. The men were asking her why her husband was not standing instead of her. And she is brilliant.Neighbouring constituency to us. One of my clients said that she didnt know who her household would be voting for, as dad hadnt made his mind up yet. He had fallen out with Tony Blair though. I really do not know how Asian women as a group can be empowered. As a party, I'm afraid we have to plug away at the menfolk, as quite a few marginals are slanted in a way that makes the Muslim vote crucial. That is, across West yorkshire for starters.

[Yet Another Anon]

>>>>Any one suggest why the Lib Dems are so high (in my view) at 20% in the polls?<<<<
Just continuing resentment by many traditional Labour supporters at the 2003 War in Iraq along with the fact that they have for the first time in several years got a leader who is focused with regard to policy direction and the hullabaloo of the leadership campaign is over.

It's also heading towards mid-term, i have no doubt that they will probably have some more by-election successes and some good Local election Results but that towards the next General Election it will come down to being between Labour and the Conservatives, and the Liberal Democrats will lose everything they have gained since 1992.

[Yet]

>>>>Internet polls at the last election showed Labour as crushed with around 22% against Lib Dem 23% with Cons on 35%. Internet polling samples were vast.<<<<
A lot of people still don't have access to computers and there are always people who vote on multiple occassions, poorer people are more likely to have quite brief access to the internet and it's more likely to be dialup narrowband so they are less likely to vote and while the difference between the voting tendencies of people on different incomes may have narrowed in the past 25 years, Labour Party support is still strongest amongst the poorest parts of society.

[Yet Another Anon]

>>>>On Mrs T - in 1983 MORI had the Tories at 46%, when they actually got 44%. In 1987 they had the Tories at 44% when they actually got 43%<<<<
Actually the Conservatives got 42.4% of the vote in 1983 and 43.7% of the vote in 1987, in fact their total vote only declined in 1983 and otherwise total numbers of votes increased in both the 1987 and 1992 General Elections, in fact they won the 1992 General Election despite the fact that as a percentage of the total possible vote it was actually Labour's best result since 1970 - turnout can make a big difference, Labour's biggest danger is sheer apathy by Labour voters who might think the General Election is in the bag and not bother voting.

[Yet Another Anon]

>>>>In the last two of those, Labour's victory was so emphatic that nobody but the pollsters noticed that their lead was exaggerated.<<<<
Most sensible people including many pollsters actually thought it would be even closer tham it eventually was, in fact I even recall Claire Short (not that she's sensible exactly the opposite but she's someone who couldn't spin a coin) saying that things would narrow and it would end up like the 1964 General election, many others thought that Labour's majority would only be around 100, I didn't seriously expect the Conservative vote to drop below 35% in 1997 and I thought the 2001 General Election would be closer than it was as well.

Actually in the last General Election from a couple of years out I thought that Labour's vote would be a bit higher than it was (37% whereas they actually got 35.2% of the vote) eventually and the Liberal Democrat vote rather lower (About 21% compared to the 22.3% they actually got) but was very close in guessing what the Conservative percentage vote would be (33% compared to 32.7% in the UK as a whole that they eventually got)) although I thought there would be a slightly higher turnout (About 67%). I rather expected a Labour majority in the 80's, I also thought that UKIP might gain a couple of seats).

[R UK]

I agree Yet. But what interested me was that they got the Conservative share and the Lib Dem share both right.

It was only Labour's share that was completely wrong.

It may be that once you get a computer you don't vote Labour. Or it may be that postal votes organised by Labour were not genuine. Not proof either way, but interesting.

[Peter Harrison]

R Uk - Any form of self selecting poll is guaranteed to be inaccurate, especially where respondents are allowed to vote multiple times. For a poll to be accurate you need to ensure that respondents can only vote once and that the results are weighted demographically to match the population at large. Your suggestion of a conspiracy involving all the polling companies is, quite frankly, unbelievable.

And by the way, the reason you don't hear any more about John Humphrys "stolen" vote is that he admitted to the Daily Telegraph (that well known left leaning newspaper) that he had turned up at the wrong polling station with his postal vote in his hand. This admission came BEFORE the bullying you allege. His vote was not stolen by anybody. It is true that he was unable to cast his vote and there is a dispute as to what advice he was given by officials at the polling stations. He claims he was told by the officials at the first polling station that the postal voting form was of no use so he destroyed it. The officials dispute this. Both he and they agree that, had he turned up at the correct polling station with postal vote in hand, he would have been able to vote.

[R UK]

I agree with all you say. My interpretation is different.

John Humphreys' story is still odd. Is it likely that he made so many errors? And then went national broadcasting what was obviously a cock-up? He is not usually known for craeting monummental messes.

I think his first reaction/story that his vote had been stolen was the truth, and his later rather complicated strange explanation which you have faithfully regurgitated, was a more politic considered version, - one which interested parties would agree to.

I think that professional journalists owe us an explanation as to how many of the 6 million postal votes went to which party. We have never been told. Also what percentage of those were definitely not fraudulent? Leaving a figure of doubtful postal votes to consider.

Why has this crucially important information never even been mentioned by any media? Do you blame me for being a tad suspicious?

[Yet Another Anon]

>>>>I agree Yet. But what interested me was that they got the Conservative share and the Lib Dem share both right.<<<<
The fact that it was roughly around that that eventually happened is coincidence, many of the votes on the internet poll referred to obviously were undecided so in theory if that had been true the Conservative Party would have got much more than 35% of the vote and the Liberal Democrats would have got more than 23% of the vote, in fact the total UK vote for the Conservative Party in the election was 32.7% of the vote which is even lower than those saying that they were going definitely to vote Conservative in the Internet Poll and the Liberal Democrats final vote was only 22.3%.

In addition many voting on it may not have been registered voters, with many being from outside the UK.

[R UK]

How do you know it was coincidence?

Maybe the figures 'might have been coincidence' but also they might have been as they were because they are a lot more accurate than you think.

Returning to other more tangible evidence, Murdoch was persuaded by his pre-election polling to consider backing Howard. See press two weeks previous to poll. His calculations were based on an estimate that there 4 million postal votes - many known to be fraudulent as his journalists were claiming at the time, but in their calculations at 4 million, not enough to make up for the fact that many labour voters were not going to vote, or were going to vote elsewhere.

It was only when it dawned on Murdoch that although Labour were going to trounced at the polling booths - (possibly in the proportions as per One Big Vote) - they had more than made up for that with enough postal votes (however acquired), that he realised he had to swing back to get behind Blair after all...if he wished to back the winner in traditional style.

If you read the papaers in the last two weeks prior to the election you can see his manoevrings quite well, and you can see the 4 million estimated number of postal votes frequently quoted, until all mention of the subject suddenly stopped, and he appeared to switch sides.

There must have been a moment of realisation that the 4 million figure although in itself shocking enough, was nothing near the actual figure, which was in fact 6 million. This can be found quoted in the Spectator inter alia.

As the gap between Labour and Conservative was less than 1 million (9.4 million v 8.6 million), it is easy to see that the postal vote numbers were more than big enough to affect the outcome.

Especially when you look at the fact that a mere 100,000 votes in key marginals can swing the result.

Postal vote fraud remains the non-subject of our age. We are British you know!!!!
No one, it seems dares to ask. The repercussions to a journalist's career prospects for asking about such delicate matters are obvious.

But does no one even dare to think?

I hope that a few more start to think the unthinkable.

[Andrew Kennedy]

R UK - I am sorry to disappoint you, or debase your conspiracy theories, but I was Campaign Director in a target seat and I did not witness any Postal Vote fraud whatsoever.

I am not saying it did not happen in some isolated incidents, or the potential is not there for it to develop in future, but I have no evidence of it happening at all in my own constituency, less still in an organised way by a mainstream political party.

In fact, in my own patch, over 50% of recorded postal voters are known Conservative pledges, and a large sample taken at the "postal vote opening" indicated we had won the PV by a substantial margin.

[Yet Another Anon]

Maybe the figures 'might have been coincidence' but also they might have been as they were because they are a lot more accurate than you think.
In which case you are suggesting that a group entirely consisting of people who had found the site and voted was somehow representative of the British population generally and that all those not expressing support for one of the 3 main party's were going to vote for other party's, if that really had been a true representation of the situation then the extra fraudulent postal votes as you say there were - I don't know of any actual statistics on the actually votes in found Postal Fraud, certainly the system has severe flaws but they are just as exploitable by any group, but anyway if it had been to the extent that Labour's support was swelled from 22% or thereabouts to 35.2% then the proportions for other party's would have dropped substantially because the total numbers of another party would have increased so actually even if what you say was true it would still be a coincidence because the Conservatives and Liberal Democrats (especially the Liberal Democrats) would only have been able to get very similar results by a huge surge in the vote between then and the General Election.

[Yet Another Anon]

>>>>Maybe the figures 'might have been coincidence' but also they might have been as they were because they are a lot more accurate than you think.<<<<
That bit should have been in brackets, but the point is either way it is a coincidence.

[Peter Harrison]

I know that the figure of 6 million postal votes has been widely quoted. However, according to the Electoral Commission, the returns from the constituencies showed that the true number cast was under 4 million. About 5.4 million papers were issued, 4.1 million were returned and just under 4 million were actually included in the count.

Details here (sorry for the long link):
http://www.electoralcommission.gov.uk/files/dms/Generalelection2005-postalvotingproxiesandspoilts_18868-13891__E__N__S__W__.xls

[R UK]

Interesting that the Spectator and others had the wrong figure on postal voting.

Still plenty of opportunity for skulduggery in amongst 4 million inside a system that allows political parties to control and submit postal votes themselves....which has not been corrected since the election, and will not be corrected.

The numbers needed to swing an election are not that many.

Reassuring to hear from Andrew Kennedy.

But if fraud exists, it will not shout from the rooftops. In fact no one wants to speak publicly about dirt as a rule.

[Yet Another Anon]

>>>>But if fraud exists, it will not shout from the rooftops. In fact no one wants to speak publicly about dirt as a rule.<<<<
If such a situation resulted in an overall victory for one party then certainly the Opposition party's would claim such, in fact in such a situation Michael Howard would not have accepted defeat and would rather have launched a court challenge to the results on the grounds that the Conservative Party had rightfully won the 2005 General Election, the Respect Party claimed that they were cheated out of a couple of seats by Labour due to postal voting but the Conservative Party and Liberal Democrats accepted the election result while pointing out that it was by the lowest percentage vote for a one party government in history.

[R UK]

As soon as Conservatives mentioned postal voting Labour cunningly found Cnservatives copying their tactics and had them prosecuted. If you control the media, you can make your crime seem like it was all someone else's fault. That's what they did.

The problem is that the public could not take it in if the election had been won by deceipt. Michael Howard would have been seen as a bad loser.

As is well known amongst the evil, tell a small lie, and people often spot it. Tell a big one, and everyone will assume you could not possibly be lying. Thus it would be with fraudulent elections.

Put it the other way around. Why tolerate fraud (and intentionally maintain a system that is known to be fraudulent) unless you know it will be useful in altering outcomes of elections?

[Yet Another Anon]

>>>>The problem is that the public could not take it in if the election had been won by deceipt. Michael Howard would have been seen as a bad loser.<<<<
If things really did get to the state that you are suggesting then the opposition would not accept defeat because that would just leave things at the risk of repeat situations in future.

If a Court Challenge failed then things would move onto more direct means such as mass demonstrations and if that failed a military coup.

But the simple fact is that Conservative Support was little changed on 2001 and may have been the very slightly bit higher because of higher turnout but was still well down on 1997 and so Labour won the General Election by default in the situation of mass apathy and a fragmented opposition vote despite their 2nd lowest vote as a percentage of those eligible to vote since 1918.

[R UK]

I am not so sure